1999 V70XC Plan of attack.

[Ozark Lee]

I have had an ongoing, somewhat random, no start issue with my new 1999 XC. It acts like a bad antenna ring but I replaced the ring and I have used DICE to scan the car. It did have an IMM-321 error which I cleared but it continues to give me grief. When it does start it starts fine and runs just fine, but it usually only starts about on time in five. The IMM-321 doesn't come back. The new car came only with one key and no remote locking keyfobs.

I bought the car to replace my other 1999 XC that got wrecked and I bought it in a non running condition due to a bad fuel pump. I had just replaced the fuel pump on the XC that got wrecked about a month before the wreck so I pulled it back out and replaced the pump on the new car with the barely used pump from the parts car. The pump runs fine and it has good fuel pressure. I have the wrecked parts car parked up behind my mother's barn.

Upon investigating I found that the ECM cage had been taken apart at some point and there is a ground wire spliced under the ECM connector that goes to a bolt on the fender. The wire it is spliced to is a ground wire so I'm not sure what the issue was.

I went up to the parts car today and removed the ECM, TCM, ETM, VGLA module, and the IMM module. Obviously, with the software loading I can't just swap modules out individually but I am thinking I can experiment by changing out interrelated systems but I need to devise a plan of attack.

I think the first thing I need to do is read all of the programming with DICE and save the file so that I can re-load modules as necessary. I'm not sure how to save the file(s) with VIDA/DICE.

My though was to then swap the ECM and the IMM from the parts car into the new car. Unplug the antenna ring and then plug in another antenna ring - that isn't around the key - and tape my parts car key to the antenna ring and then try to start the car. The IMM and the ECM should, in my mind, be able to communicate with each other since they are married modules. The ETM and the VGLA would not match but I'm not sure if they need to. Both cars have exactly the same engine so I think the ETM will work anyway even though it isn't specifically married to the car.

If that works and the car starts reliably I may go ahead and swap the VGLA module (although it is a PITA) in hopes that my old keyfob will work with the new car.

Are there any problems with the plan?

...Lee

[scot850]

Before going crazy on the hard stuff, I know you are one of the most experienced guys on the forum, but have you tried some of the other issues first?

1) My worst curse was the worn out ignition switch. Had a problem with inconsistent connections in the ignition barrel which caused in-consistent starts.

2) Bad chip in the key?

3) Park sensor or gear shift not going fully home?

I apologise if you have already done this, but sometimes the more obvious get overlooked. Both my 98 XC (without the proximity chip) and my 2000 (with chip) had the same issue within 6 months.

Neil.

[Ozark Lee]

I might be over-thinking it, the software aspects of this just frustrate the hell out of me. I haven't had any previous module problems so I don't have a great feel for how things behave when there is a software glitch. I really can't stand the thought of lining Rick Henrdick's bulging pockets anymore, they would likely charge me more than I paid for the car in the first place to figure it all out.

It would be easy enough to run back up to the parts car and snag the ignition switch and give that a shot first. The bad ignition switch usually presents a bunch of other symptoms but I must confess, I've only driven the car twice and it has been for very short distances so I suppose I don't really have a good feel for any of the other issues that normally present.

The key chip seems to be OK, when I scan it with DICE it doesn't show any immobilzer errors any more - just the initial error that I cleared with DICE and it hasn't come back.

The car does crank fine so I don't think the shifter is causing me problems.

...Lee

[precopster]

Hi Ozark, Vida/Dice can log all your module's version numbers before you embark on any swapping exercise. Go into Network Tab after pressing Communication and click on each module's icon to obtain hardware and software version numbers. Write these down for each module or preferably perform a screen capture (with PrtScn key) and save them to Microsoft Paint as a JPEG file.

If you swap in another module restart Vida and Choose Update before Vida does the preliminary scan of the vehicle as it will retain old Network info if you don't do this. Most likely the ETM sw version numbers are shared but even they can vary between cars of the same year and model. You can perform the start of an ETM test (without doing the ETM test or restarting Vida) and this will reveal sw numbers right away. Just hotplug the 2nd ETM and restart the ETM test.

Contact Henry or "doublebug" as he's known by PM or Skype and he will contact you regarding the next steps.

[Ozark Lee]

Well - that was an epic failure.

I really need to get a handle on the CAN bus architecture. I was under the impression that the ECM held the car's basic identify. Not so.

I first swapped the ECM and DICE continued to read the correct VIN from the cars as well as the correct mileage. The replacement ECM failed to communicate with the immobilizer module which really didn't come as a shock so I swapped the immobilizer and things continued to fail to communicate, it just got worse. Even plugging in the antenna ring away from the ignition cylinder and taping the key that was mated to that immobilizer failed to get it to read the key and the immobilizer continued to fail to communicate with the ECM.

Apparently there is a central place where the basic identity of the car is stored and through which the module characteristics are coordinated. I can't see such a module with DICE.

After screwing with it for several hours I managed to drain the battery to a point where DICE started to error on reading multiple modules so I hung a battery charger on it and buttoned things back up with the original modules and DICE telling me everything is fine which is great if the damn car would start twice in a row.

At least I got the ECC temp sensor cleaned out and working again. I swear, I am going to hear the door chime in my sleep tonight.

...Lee

[precopster]

The extraction of firmware from ECM to ECM is possible ONLY at dealer by using a particular sw part number which was released last year. Unfortunately it was only released for the P2 series.

Henry has cloned one of my customer's ECMs on an '01 S60T5 using my Dice and his own extraction software by remote connection 3 weeks ago. The customer's car started right up after the cloning procedure. Also he has extracted about a dozen ETM firmware for me and rewritten either new or used ETMs.
ARD's Lucky can do the same thing as well as XeModex as you may know by having both ECMs sent in to them.

The identity is not "held" in one place. It is a matching process of an identity key from module to module.

If you want to swap the IMM from one car to another you'll need the antenna ring, the TCM, CCM, PSM, SRS and possibly other modules (ETM will come up with a config fault upon swapping though it may not run the car properly if the sw part number doesn't match) ETM is only module not actually "married" to VIN; it just requires correct firmware to match the ECM version.

[Ozark Lee]

I started to read up on the J2534 CAN bus last night but I wound up with a bunch of IEEE and SAE protocol papers that gave me a headache. What I did gather from what little I read was that installing software is akin to installing a new operating system on a PC, complete with a Hardware Abstraction Layer so it truly is a system approach rather than being a system of discreet components that are being coordinated.

What sucks is having a beautiful, low mileage car sitting in the driveway that is like a bricked iPhone. It looks real nice but it doesn't do anything. At least I have some mental down time to think about it today as I grout tile.

...Lee

[precopster]

If you can understand this thread you can probably understand how to write some firmware extraction software: http://forums.swedespeed.com/showthread.php?t=215698

[Ozark Lee]

Thanks Mike,

That is actually a very good thread and it deals with things on terms I can get my head around, specifically physical stuff like EEPROMS. I pulled my electrical engineering degree in the mid 1980s and I never followed a career path in microprocessor based systems although I have tried to keep up with stuff that is specific to my specialty.

The whole CAN bus concept is actually kinda cool from a technological perspective but the lack of an ability to deal with things on a component level is what has me frustrated. I did find a very good summary of the original Bosch system and it has been a real help in figuring out what is going on. I don't think that this article is totally accurate for the Volvo system but I think the concepts are the same. The bit length of the headers, etc may not be accurate and I think there are additional fields that have information like the VIN number in them.

The article is here if anyone is curious:

http://www6.in.tum.de/pub/Main/Teaching ... CANbus.pdf

The key nugget there is that the CAN bus is not address driven but rather each module in the system has a unique identifier coded to it. If that unique identifier is not present in the data stream the module will ignore the subsequent data which keeps the bus traffic relatively low when compared to an address driven network which is sort of like the floor of the Stock Exchange where everyone is screaming their IP address at the same time trying to get noticed. It appears that the identifiers are handed out as a part of loading the software in the first place and they can get reassigned as changes are made to the system but it also explains how a "cloned" module would work just as well as the original since the it would retain the original identifier code.

Alas, this really doesn't help me get my car running without costly intervention but it does get me mentally prepared for biting the bullet and paying to get software reloaded. Before I go that far I am going to go back to the basics and do a more thorough test of basic stuff like cam and crank position sensors, etc. I can at least look at the signals with basic test equipment like a meter and an oscilloscope and make sure that the inputs to the system are good.

...Lee

[abscate]

Lee...someone just solved an intermitant no-start threat finding a loose ground in the upper CEM module ( up in the mirror assembly?)

fWIW